Logic cleanup & add preprocessing CI test

Author: ChAoSUnItY

.github/workflows/main.yml

@@ -56,6 +56,44 @@ jobs:
           make check-sanitizer || exit 1
           make check || exit 1
 
+  preprocessor-host:
+    runs-on: ubuntu-24.04
+    strategy:
+      matrix:
+        compiler: [gcc, clang]
+        architecture: [arm, riscv]
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+      - name: Download dependencies
+        run: |
+          sudo apt-get update -q -y
+          sudo apt-get install -q -y graphviz jq
+          sudo apt-get install -q -y qemu-user
+          sudo apt-get install -q -y build-essential
+      - name: Configurate config
+        env:
+          CC: ${{ matrix.compiler }}
+        run: |
+          make distclean config ARCH=${{ matrix.architecture }}
+      - name: Preprocess stage 1 source code
+        run: |
+          make out/shecc
+          ./out/shecc -E src/main.c > ./out/out.c
+      - name: Build stage 1 artifact
+        run: |
+          ./out/shecc --no-libc -o out/shecc-stage1.elf ./out/out.c
+      - name: Preprocess stage 2 source code
+        run: |
+          make out/shecc
+          ./out/shecc-stage1.elf -E src/main.c > ./out/out.c
+      - name: Build stage 2 artifact
+        run: |
+          ./out/shecc-stage1.elf --no-libc -o out/shecc-stage2.elf ./out/out.c
+      - name: Test stage 2 artifact
+        run: |
+          make check-stage2 || exit 1
+
   coding-style:
     runs-on: ubuntu-24.04
     steps:

src/defs.h

@@ -56,6 +56,7 @@
 #define SMALL_ARENA_SIZE 65536    /* 64 KiB - for small allocations */
 #define LARGE_ARENA_SIZE 524288   /* 512 KiB - for instruction arena */
 #define DEFAULT_FUNCS_SIZE 64
+#define DEFAULT_SRC_FILE_COUNT 8
 
 /* Arena compaction bitmask flags for selective memory reclamation */
 #define COMPACT_ARENA_BLOCK 0x01   /* BLOCK_ARENA - variables/blocks */

src/globals.c

@@ -18,6 +18,9 @@ char *intern_string(char *str);
 
 /* Lexer */
 token_t *cur_token;
+/* TOKEN_CACHE maps filename to the corresponding computed token stream */
+hashmap_t *TOKEN_CACHE;
+strbuf_t *LIBC_SRC;
 
 /* Global objects */
 
@@ -71,8 +74,6 @@ int elf_offset = 0;
 
 regfile_t REGS[REG_CNT];
 
-strbuf_t *SOURCE;
-
 hashmap_t *INCLUSION_MAP;
 
 /* ELF sections */
@@ -1223,7 +1224,8 @@ void global_init(void)
         arena_alloc(GENERAL_ARENA, sizeof(string_literal_pool_t));
     string_literal_pool->literals = hashmap_create(256);
 
-    SRC_FILE_MAP = hashmap_create(8);
+    TOKEN_CACHE = hashmap_create(DEFAULT_SRC_FILE_COUNT);
+    SRC_FILE_MAP = hashmap_create(DEFAULT_SRC_FILE_COUNT);
     FUNC_MAP = hashmap_create(DEFAULT_FUNCS_SIZE);
     CONSTANTS_MAP = hashmap_create(MAX_CONSTANTS);
 
@@ -1351,7 +1353,7 @@ void global_release(void)
     arena_free(TOKEN_ARENA);
     arena_free(GENERAL_ARENA); /* free TYPES and PH2_IR_FLATTEN */
 
-    strbuf_free(SOURCE);
+    strbuf_free(LIBC_SRC);
     strbuf_free(elf_code);
     strbuf_free(elf_data);
     strbuf_free(elf_rodata);
@@ -1360,6 +1362,7 @@ void global_release(void)
     strbuf_free(elf_strtab);
     strbuf_free(elf_section);
 
+    hashmap_free(TOKEN_CACHE);
     hashmap_free(SRC_FILE_MAP);
     hashmap_free(FUNC_MAP);
     hashmap_free(CONSTANTS_MAP);
@@ -1694,7 +1697,7 @@ void error_at(char *msg, source_location_t *loc)
 /* FIXME: This function has been deprecated, use `error_at` instead. */
 void error(char *msg)
 {
-    printf("[Error]: %s\nOccurs at source location %d.\n", msg, SOURCE->size);
+    printf("[Error]: %s.\n");
     abort();
 }
 

src/lexer.c

@@ -27,10 +27,6 @@ hashmap_t *KEYWORD_MAP = NULL;
 token_kind_t *directive_tokens_storage = NULL;
 token_kind_t *keyword_tokens_storage = NULL;
 
-/* TOKEN_CACHE maps filename to the corresponding computed token stream */
-hashmap_t *TOKEN_CACHE = NULL;
-strbuf_t *LIBC_SRC;
-
 void lex_init_directives()
 {
     if (DIRECTIVE_MAP)
@@ -934,9 +930,6 @@ token_stream_t *lex_token_by_file(char *filename)
     source_location_t loc = {0, 1, 1, 1, filename};
     strbuf_t *buf;
 
-    if (!TOKEN_CACHE)
-        TOKEN_CACHE = hashmap_create(8);
-
     tks = hashmap_get(TOKEN_CACHE, filename);
 
     /* Already cached, just return the computed token stream */
@@ -1000,9 +993,6 @@ token_stream_t *include_libc()
     if (tks)
         return tks;
 
-    if (!TOKEN_CACHE)
-        TOKEN_CACHE = hashmap_create(8);
-
     if (!hashmap_contains(SRC_FILE_MAP, filename))
         hashmap_put(SRC_FILE_MAP, filename, LIBC_SRC);